Students of Old English metrics are well aware of the importance of grammar for understanding the nature of metre. The systematic analyses of Sievers, Pope and Bliss, however much they differ in emphasis and detail from each other, are based on careful investigations of the patterns of stress and syntax revealed by Old English verse. Their work has established beyond a doubt the general validity of the metrical classifications which are usually identified by the Sievers types (Bliss's typology, which is derived from Sievers's, has been gaining in acceptance), even though the significance of the classifications, especially for performance, remains very much an open question. Nevertheless the Sievers typology is limited in its usefulness by its rather too narrow focus on metre, which has tended to mask some of the evidence of the metrical grammar.

The study of the metrical grammar which is the subject of this book puts this evidence in a new light. The level of abstraction represented by the five types is misleading when, for example, it leads to the quite valid generalization that the ‘types may occur in any combination or sequence’. When properly analysed, the metrical types are not completely interchangeable. And I do not refer merely to the well-known fact that double alliteration limits any half-line to the a-verse.

The onset of the half-line determines which of the three basic types it belongs to.